Re: seeking clarification on the cladism debate (RE: hidden "cladistic" ranks)

[Graydon <graydon@dsl.ca>]

On Thu, Oct 18, 2001 at 04:18:39AM +0000, Ken Kinman scripsit:
> Thus unlike traditional "eclectic" classifications, paraphyletic
> groups are clearly marked as such, and sister group information is
> retained. 

Well, Ken, speaking strictly as an interested amateur, what you
described in the section I snipped above is exceedingly confusing.  I
can rhyme off 'kingdom phylum class order family genus species' and I
have a decent grasp of where the order subsets and such go, but if you
start shifting stuff around it disolves very quickly into mush.

The arguments for hanging on to genus as a designation, particularly in
paleontology, strike me as very strong; I have yet to see any argument
for hanging on to the Linnean system, something invented before Darwin's
formulation of Natural Selection!

> And it does this without the adverse "side-effects" of purely
> cladistic classifications (e.g., multiplicity of names, hierarchical
> instability, lack of anagenetic information, and decreases in
> utility).  

Pure cladistic classifications have two really important features.

One is that they explicitly represent a specific understanding of the
evolutionary history of a group of organisms, so that when the hierarchy
changes it represents a change of understanding.  This is a built in
acknowledgment of the magnitude of ignorance, too, and I really like
that part, and the corresponding tendency to associate dates with trees.

I'm not a specialist; I'm just curious (and think dinosaurs are Really
Neat.)  Arguments about what's an Order and what's a class make very
little sense; cladograms say 'we think the history of the breeding
population of things descended from this organism here looked like', and
that's conceptually easy to follow.

Two is that they're really, really simple; the cladograms being
presented on this list are simple binary tree structures.  They're much
easier to understand that the distinction between the subtypes of
'Order' level classifications.

That the nature of a character and how they're coded is controversial is
also a good thing; there's some really fearsome stats lurking under the
question of 'what's meaningful osteological variation?', and having a
system which explicitly demands statistical analysis on that point seems
to me a very good thing.

Science works by admitting one is wrong, not by getting things right, and
cladistics seems to be an approach that makes admitting that one is
wrong a commonplace.

Since the only stable tree is the one which represents what actually
happened, rather than the -- necessarily imperfect, fragmentary, and
forever missing critical data -- human understanding of what happened,
having a classification system which will shift as a dynamic thing
representing a dynamic understanding is a net gain.

-- 
                           graydon@dsl.ca
               To maintain the end is to uphold the means.